BTC $79,600.16 -0.26%
ETH $2,489.89 -1.73%
BNB $706.56 -0.63%
XRP $1.42 -1.02%
SOL $106.24 +1.63%
TRX $0.3392 +0.88%
DOGE $0.0874 -1.96%
ADA $0.2092 -2.83%
BCH $261.91 -4.23%
LINK $11.66 -1.59%
HYPE $83.43 +1.42%
AAVE $125.30 -2.87%
SUI $0.7598 -1.75%
XLM $0.1836 -1.96%
ZEC $788.62 -0.45%
BTC $79,600.16 -0.26%
ETH $2,489.89 -1.73%
BNB $706.56 -0.63%
XRP $1.42 -1.02%
SOL $106.24 +1.63%
TRX $0.3392 +0.88%
DOGE $0.0874 -1.96%
ADA $0.2092 -2.83%
BCH $261.91 -4.23%
LINK $11.66 -1.59%
HYPE $83.43 +1.42%
AAVE $125.30 -2.87%
SUI $0.7598 -1.75%
XLM $0.1836 -1.96%
ZEC $788.62 -0.45%

ssi

All
Article
Flash

first_img SemiAnalysis: HBF non-HBM alternative, cost and heat dissipation still have uncertainties

P Equity Research and SemiAnalysis researcher Nick Doyle and others discussed high bandwidth flash (HBF) in X Space. Nick stated that it is still too early to determine how much the cost premium of HBF relative to HBM can shrink; existing data mostly comes from vendor claims, such as Sandisk stating that the cost per bit is about one-eighth that of HBM. Yields, testing, and other factors will improve with scale, but structural costs such as TSV, stacking, and pSLC mode will always exist, and durability is a key unknown; if wear exceeds expectations, costs will rise.The application scenarios for HBF are narrow, targeting only AI inference, especially low batch and long context MoE models, and it is not a substitute for HBM. The actual bandwidth target is about 1.6 TB/s, which is at the HBM3E level, suitable for sequential reads to load model weights, more aligned with the capacity needs of a small number of GPUs in local or private enterprises, rather than ultra-large-scale bandwidth scenarios. Heat dissipation reliability has not yet been resolved; flash memory will degrade faster at high temperatures next to GPUs, and mitigation measures such as UCIe separation and daily refresh have yet to be validated.In terms of manufacturing, Sandisk/Kioxia has experience with 3D NAND, and SK Hynix complements HBM-style stacking capabilities, but mass production is still to be confirmed. Overall, storage is shifting towards a specialized layered market, with NAND shortages expected to continue until 2028, and HBF may further impact supply and demand.

first_img Hyperliquid's policy center suggests to the SEC and CFTC to classify perpetual equity as securities futures

Hyperliquid Policy Center (HPC) submitted a letter of opinion to the U.S. Securities and Exchange Commission (SEC) and the Commodity Futures Trading Commission (CFTC) stating that eligible equity perpetual contracts can be classified as securities futures. This category is jointly regulated by the two agencies, allowing exchanges to compete on execution quality rather than jurisdiction.HPC pointed out that there is still no clear classification for perpetual contracts under U.S. law (whether futures or swaps), and this fundamental issue remains unresolved. They possess characteristics of futures such as standardized terms, the ability to hedge positions, and forward value, although they do not have a fixed expiration date, but prices converge continuously through funding rates.The letter of opinion proposed four points: confirm that the definition of securities futures can encompass cash-settled equity perpetuals with futures characteristics; retain flexibility for product listings at trading venues; maintain consistency in classification between the two agencies; modernize the securities futures framework to accommodate new structures. HPC stated that in the past 10 months, the trading volume of Hyperliquid perpetual contracts has exceeded $48 billion, and a clear framework would help relevant products enter the U.S. market.

first_img Goldman Sachs: Nvidia's impressive earnings report may struggle to boost stock prices

According to the Financial Associated Press, Nvidia will announce its Q2 2027 financial report after the US stock market closes on August 26. Goldman Sachs expects strong performance with potential upward revisions in guidance, but notes that the stock price has already surged since August, with an increase of over 12% in the past two weeks, suggesting that the good news may have already been priced in, making it difficult for excellent performance to boost the stock price. Goldman Sachs' EPS forecasts for Nvidia in Q2 and Q3 are 6% and 12% higher than Wall Street's consensus expectations, respectively, with a target price of $285, which is still slightly below the market average.Goldman Sachs believes there is room for revaluation in the valuation, provided that the profitability of large-scale cloud vendors improves to support capital expenditures, Nvidia maintains caution on its customer financing platform to alleviate concerns about supplier circular financing, and continues to advance large-scale buybacks and dividends. Investors will focus on details of the customer financing platform in the earnings call, progress on the Vera Rubin platform, gross margins and raw material costs, CPU demand driven by AI agents, and the competitive landscape in the industry.Goldman Sachs points out that while GPU demand is strong, there is still a risk of a pullback, and the difficulty of "exceeding expectations and raising guidance" increases each quarter, raising doubts about whether cloud vendors can continue to ramp up capital expenditures. Earlier this month, Nvidia, in collaboration with Apollo, BlackRock, Blackstone, Goldman Sachs, KKR, and others, established a financing platform for AI computing infrastructure, aiming to leverage over $500 billion in third-party capital; concerns about circular financing are rising in the market, and the company needs to release more substantial new information, or else the stock price is more likely to decline rather than surge after the earnings report.
app_icon
ChainCatcher Building the Web3 world with innovations.